A Crosslinguistic Study of Reference and Cognitive Status

University Of Minnesota-Twin Cities, Minneapolis MN

Investigators

Abstract

A characteristic feature of human language is that different forms can be used to refer to the same thing and the same form can refer to different things; yet people still understand one another. For example, the appointment of a particular individual to be US ambassador to the United Nations can be referred to as 'this appointment', 'the appointment', 'it', 'this', or 'that'; and each of these forms could refer to other entities on different occasions of use. A partial explanation for such facts is that words such as 'a', 'the', 'this', 'that' and 'it' constrain possible interpretations by encoding information about the hearer or reader's assumed memory and attention state (cognitive status) in relation to the entity referred to, for example whether he or she already has a mental representation of the referent (familiar) and whether it is the focus of attention. While all languages have words that encode cognitive status, they may differ in the range and number of words that have this function. For example, some languages (e.g. Japanese and Russian) have no articles like English 'the' or 'a', while others (e.g. Arabic and Norwegian) have double marking of definiteness, resulting in forms roughly equivalent, for example, to 'the this appointment'. Languages also differ in the statuses coded by corresponding forms. For example, the demonstrative determiner 'this' in English, as in 'this hat', requires the intended referent to be in the addressee's short term memory (activated), while the corresponding form in Russian only assumes familiarity with the referent. With support from the National Science Foundation, Dr. Jeanette Gundel will study the relation between referring forms and cognitive status, extending the research to languages that have not yet been studied in this respect -Tunisian Arabic, Norwegian, Kumyk (a Turkic Language spoken in Southern Russia), and Ojibwe (an Algonquian language). In addition to furthering our knowledge of how people produce and understand referring forms and the extent of similarity and difference across languages, this project will contribute to our understanding of how knowledge of the language system interacts with other aspects of cognition. A characteristic assumption of the theory being tested that distinguishes it from more traditional accounts is that forms are more or less specified for different statuses rather than being in opposition. This explains why the same entity can sometimes be referred to in different ways even in the same context. For example, 'the', which requires the ability to associate a unique referent, is more specified for cognitive status, and therefore more restrictive, than 'a', which only requires identification of the type; and 'it', which requires the referent to be in focus of attention, is more specified than 'this', which only requires the referent to be in short term memory. A major component of the project is development of explicit procedures for determining the assumed cognitive status of a referent. The project thus also has potential applications in such areas as machine translation, computer generation and understanding of human language, and automated information extraction and retrieval from texts. The methodology will involve analysis of texts from diverse genres, both spoken and written as well elicited native speaker judgments of constructed examples. The project will provide opportunities for graduate and undergraduate students to engage in research.